Posted by daestrom on February 12, 2009, 6:40 pm
David Williams wrote:
> On 02/12/2009 4:44 AM, tim@tim-jackson.co.uk wrote to All:
> -> I'm not sure how much we need to measure current anyway. We *know*
> how
> -> much current a 60W light bulb draws, it's 260mA here.
> Probably not! The temperature of a filament of an incandescent lamp
> running off 110- or 230-volt AC varies considerably during an AC
> half-cycle. When it's relatively cool, its resistance is
> substantially lower than when it's hot. As a result, the current is
> not sinusoidal. The peak current is displaced toward the start of the
> half-cycle, but the zeroes of the current and voltage coincide. Even
> stranger things happen if the lamp is dimmed with a triac-type
> dimmer. The maximum curent *increases* as the lamp is dimmed,
> provided it is not dimmed very much. The filament is cool when the
> maximum voltage hits, so
> the current is large.
> The average (volts x amps) of a 60-watt bulb is 60^2. But the RMS
> current is not the quotient of the power and the RMS volts.
I'm not sure I buy into that. Does the filament temperature *really* swing
that fast? I would think the mass and heat capacitance 'evens out' the
temperature swings and the filament stays hot as voltage passes through
zero.
After all, the filament continues to glow for a moment after switching off,
especially in larger bulbs.
If it cooled as much as you imply, there would be stroboscopic 'flicker'
from incandescent bulbs, similar to that of many flourescent ones. Don't
see that.
daestrom
Posted by Johnny B Good on February 12, 2009, 9:43 pm
> David Williams wrote:
> > On 02/12/2009 4:44 AM, tim@tim-jackson.co.uk wrote to All:
> >
> > -> I'm not sure how much we need to measure current anyway. We *know*
> > how
> > -> much current a 60W light bulb draws, it's 260mA here.
> >
> > Probably not! The temperature of a filament of an incandescent lamp
> > running off 110- or 230-volt AC varies considerably during an AC
> > half-cycle. When it's relatively cool, its resistance is
> > substantially lower than when it's hot. As a result, the current is
> > not sinusoidal. The peak current is displaced toward the start of the
> > half-cycle, but the zeroes of the current and voltage coincide. Even
> > stranger things happen if the lamp is dimmed with a triac-type
> > dimmer. The maximum curent *increases* as the lamp is dimmed,
> > provided it is not dimmed very much. The filament is cool when the
> > maximum voltage hits, so
> > the current is large.
> >
> > The average (volts x amps) of a 60-watt bulb is 60^2. But the RMS
> > current is not the quotient of the power and the RMS volts.
> >
> I'm not sure I buy into that. Does the filament temperature *really* swing
> that fast? I would think the mass and heat capacitance 'evens out' the
> temperature swings and the filament stays hot as voltage passes through
> zero.
The effect is real, if perhaps overstated.
> After all, the filament continues to glow for a moment after switching off,
> especially in larger bulbs.
Yes, thermal inertia will moderate the effect somewhat but that doesn't
change the basic truth of the matter, even though the effect is, in
practice, an insignificant one.
> If it cooled as much as you imply, there would be stroboscopic 'flicker'
> from incandescent bulbs, similar to that of many flourescent ones. Don't
> see that.
You could try the optical telephone link experiment which made use of a
torch (flashlight for you goddamn yanks;-) and a half watt audio amp for
the transmitter and a CdS cell and amplifier for the reciever.
Essentially, the 4 ohm output winding was wired in series with the 3
volt half amp lamp typically used in torches of the day (60's, 70's?)
allowing the brightness to be modulated by the output of the audio amp.
You could see the brightness vary with the modulation and, wonder of
wonders, hear a rather muffled version in the optical reciever. The
brighter you ran the lamp, the less muffled it would be (up to a point!
;-).
The effective 'depth of modulation' was probably just a few percent,
but it was enough to produce a 'result'. Clearly, in the case of the
Yankee system, a GLS tungsten filament lamp will be producing a rather
weak 120Hz flicker component which will be invisible for two reasons.
Firstly, and foremost, because the 120Hz flicker rate is well above what
the human visual system can percieve, and, secondly, because the
effective depth of modulation is probably less than 5%.
If we compare this with the 90% depth of modulation flicker on a CRT
based monitor running a non interlaced scan at a vertical refresh rate
of 75Hz (which only a tiny fraction of the population can percieve) or
100Hz refresh rate (which none of the population can percieve), you'll
understand why you don't percieve the 120Hz 'flicker' from a GLS
tungsten filament lamp.
However, a fluorescent tube, rather like the CRT monitor has a
significantly strong component, mostly at 120Hz with a lower level at
60Hz due to a partial rectification effect (normally negligable in new
tubes) which becomes more pronounced as it approaches the end of its
service life.
Once a fluorescent tube develops the first hint of flicker, it's
usually a reliable sign that it has completed its 6 to 8 thousand hours
'tour of duty' and needs to be replaced.
HTH
--
Regards, John.
Please remove the "ohggcyht" before replying.
The address has been munged to reject Spam-bots.
Posted by Tim Jackson on February 13, 2009, 2:53 am
daestrom wrote:
> David Williams wrote:
>> On 02/12/2009 4:44 AM, tim@tim-jackson.co.uk wrote to All:
>>
>> -> I'm not sure how much we need to measure current anyway. We *know*
>> how
>> -> much current a 60W light bulb draws, it's 260mA here.
>>
>> Probably not! The temperature of a filament of an incandescent lamp
>> running off 110- or 230-volt AC varies considerably during an AC
>> half-cycle. When it's relatively cool, its resistance is
>> substantially lower than when it's hot. As a result, the current is
>> not sinusoidal. The peak current is displaced toward the start of the
>> half-cycle, but the zeroes of the current and voltage coincide. Even
>> stranger things happen if the lamp is dimmed with a triac-type
>> dimmer. The maximum curent *increases* as the lamp is dimmed,
>> provided it is not dimmed very much. The filament is cool when the
>> maximum voltage hits, so
>> the current is large.
>>
>> The average (volts x amps) of a 60-watt bulb is 60^2. But the RMS
>> current is not the quotient of the power and the RMS volts.
>>
>
> I'm not sure I buy into that. Does the filament temperature *really*
> swing that fast? I would think the mass and heat capacitance 'evens
> out' the temperature swings and the filament stays hot as voltage passes
> through zero.
>
> After all, the filament continues to glow for a moment after switching
> off, especially in larger bulbs.
>
> If it cooled as much as you imply, there would be stroboscopic 'flicker'
> from incandescent bulbs, similar to that of many flourescent ones.
> Don't see that.
>
> daestrom
The effect is there, but quite small, You can see it if you shoot a
video clip under artificial light (with a battery powered camera -
web-cams can synchronise). But I think the effect of crest factor on
current is rather less than production tolerance and that produced by
supply variations.
Anyway, what was your point? We were talking about saving energy. My
point was that the power consumption of a particular bulb was
essentially *constant* and didn't need to be measured continuously, I
accept that the example number I cited was subject to some small
variations beyond the control of the user.
In practical terms the lamp is either on or off, unless of course you
have a dimmer switch, but those appear to be deprecated now, as they
require incandescent lamps, and make them even less efficient.
Tim
Posted by harry on February 13, 2009, 3:49 pm
> daestrom wrote:
> > David Williams wrote:
> >> On 02/12/2009 4:44 AM, t...@tim-jackson.co.uk wrote to All:
> >> -> I'm not sure how much we need to measure current anyway. We *know*
> >> how
> >> -> much current a 60W light bulb draws, it's 260mA here.
> >> Probably not! The temperature of a filament of an incandescent lamp
> >> running off 110- or 230-volt AC varies considerably during an AC
> >> half-cycle. When it's relatively cool, its resistance is
> >> substantially lower than when it's hot. As a result, the current is
> >> not sinusoidal. The peak current is displaced toward the start of the
> >> half-cycle, but the zeroes of the current and voltage coincide. Even
> >> stranger things happen if the lamp is dimmed with a triac-type
> >> dimmer. The maximum curent *increases* as the lamp is dimmed,
> >> provided it is not dimmed very much. The filament is cool when the
> >> maximum voltage hits, so
> >> the current is large.
> >> The average (volts x amps) of a 60-watt bulb is 60^2. But the RMS
> >> current is not the quotient of the power and the RMS volts.
> > I'm not sure I buy into that. Does the filament temperature *really*
> > swing that fast? I would think the mass and heat capacitance 'evens
> > out' the temperature swings and the filament stays hot as voltage passes
> > through zero.
> > After all, the filament continues to glow for a moment after switching
> > off, especially in larger bulbs.
> > If it cooled as much as you imply, there would be stroboscopic 'flicker'
> > from incandescent bulbs, similar to that of many flourescent ones.
> > Don't see that.
> > daestrom
> The effect is there, but quite small, You can see it if you shoot a
> video clip under artificial light (with a battery powered camera -
> web-cams can synchronise). But I think the effect of crest factor on
> current is rather less than production tolerance and that produced by
> supply variations.
> Anyway, what was your point? We were talking about saving energy. My
> point was that the power consumption of a particular bulb was
> essentially *constant* and didn't need to be measured continuously, I
> accept that the example number I cited was subject to some small
> variations beyond the control of the user.
> In practical terms the lamp is either on or off, unless of course you
> have a dimmer switch, but those appear to be deprecated now, as they
> require incandescent lamps, and make them even less efficient.
> Tim
CFls are now available that can be dimmed with a standard dimmer. Also
ones that can be dimmed by an ordinary light switch. They work quite
well too.
Years ago fluorescent lights could still be dimmed. There existed a
device that pruduced a regular high voltage spike that maintained the
arc in the tube. It worked well but was expensive and reduced the tube
life.
Posted by Jim Wilkins on February 13, 2009, 9:41 am
On Feb 12, 3:08 pm, david.willi...@bayman.org (David Williams) wrote:
> On 02/12/2009 4:44 AM, t...@tim-jackson.co.uk wrote to All:
> Probably not! The temperature of a filament of an incandescent lamp running
> off 110- or 230-volt AC varies considerably during an AC half-cycle. When it's
> relatively cool, its resistance is substantially lower than when it's hot. ..>
> dow
I've used light bulbs as dropping resistors in a wide-range Zener
regulator and measured the DC V-I curve. Basically it's logarithmic,
the current doubles for ten times the voltage because the filament
resistance increases as it heats. The effect dow mentioned depends on
the thermal mass of the filament.
If you want to see some crazy current fluctuations, look at
fluorescent tubes.
Jim Wilkins
XML site map
> -> I'm not sure how much we need to measure current anyway. We *know*
> how
> -> much current a 60W light bulb draws, it's 260mA here.
> Probably not! The temperature of a filament of an incandescent lamp
> running off 110- or 230-volt AC varies considerably during an AC
> half-cycle. When it's relatively cool, its resistance is
> substantially lower than when it's hot. As a result, the current is
> not sinusoidal. The peak current is displaced toward the start of the
> half-cycle, but the zeroes of the current and voltage coincide. Even
> stranger things happen if the lamp is dimmed with a triac-type
> dimmer. The maximum curent *increases* as the lamp is dimmed,
> provided it is not dimmed very much. The filament is cool when the
> maximum voltage hits, so
> the current is large.
> The average (volts x amps) of a 60-watt bulb is 60^2. But the RMS
> current is not the quotient of the power and the RMS volts.